Characterisation of ovine bone marrow-derived stromal ... · oBMSC micro-pellets, while another two...

RESEARCH Open Access

Characterisation of ovine bone marrow-derived stromal cells (oBMSC) andevaluation of chondrogenically inducedmicro-pellets for cartilage tissue repairin vivoK. Futrega1,2,3†, E. Music3,4†, P. G. Robey2, S. Gronthos5, R. Crawford1, S. Saifzadeh1, T. J. Klein1 andM. R. Doran1,2,3,4,6*

Abstract: Bone marrow stromal cells (BMSC) show promise in cartilage repair, and sheep are the most commonlarge animal pre-clinical model.

Objective: The objective of this study was to characterise ovine BMSC (oBMSC) in vitro, and to evaluate thecapacity of chondrogenic micro-pellets manufactured from oBMSC or ovine articular chondrocytes (oACh) to repairosteochondral defects in sheep.

Design: oBMSC were characterised for surface marker expression using flow cytometry and evaluated for tri-lineagedifferentiation capacity. oBMSC micro-pellets were manufactured in a microwell platform, and chondrogenesis wascompared at 2%, 5%, and 20% O2. The capacity of cartilage micro-pellets manufactured from oBMSC or oACh torepair osteochondral defects in adult sheep was evaluated in an 8-week pilot study.

Results: Expanded oBMSC were positive for CD44 and CD146 and negative for CD45. The common adipogenicinduction ingredient, 3-Isobutyl-1-methylxanthine (IBMX), was toxic to oBMSC, but adipogenesis could be restoredby excluding IBMX from the medium. BMSC chondrogenesis was optimal in a 2% O2 atmosphere. Micro-pelletsformed from oBMSC or oACh appeared morphologically similar, but hypertrophic genes were elevated in oBMSCmicro-pellets. While oACh micro-pellets formed cartilage-like repair tissue in sheep, oBMSC micro-pellets did not.

(Continued on next page)

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate ifchanges were made. The images or other third party material in this article are included in the article's Creative Commonslicence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commonslicence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtainpermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to thedata made available in this article, unless otherwise stated in a credit line to the data.

* Correspondence: [email protected]; [email protected]†K. Futrega and E. Music are co-first authors.1Centre for Biomedical Technologies (CBT), Queensland University ofTechnology (QUT), Brisbane, Queensland, Australia2National Institute of Dental and Craniofacial Research (NIDCR), NationalInstitutes of Health (NIH), Bethesda, Maryland, USAFull list of author information is available at the end of the article

Futrega et al. Stem Cell Research & Therapy (2021) 12:26 https://doi.org/10.1186/s13287-020-02045-3

http://crossmark.crossref.org/dialog/?doi=10.1186/s13287-020-02045-3&domain=pdfhttp://orcid.org/0000-0001-5876-4757http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/publicdomain/zero/1.0/mailto:[email protected]:[email protected]

(Continued from previous page)

Conclusion: The sensitivity of oBMSC, compared to human BMSC, to IBMX in standard adipogenic assays highlightsspecies-associated differences. Micro-pellets manufactured from oACh were more effective than micro-pelletsmanufactured from oBMSC in the repair of osteochondral defects in sheep. While oBMSC can be driven to formcartilage-like tissue in vitro, the effective use of these cells in cartilage repair will depend on the successfulmitigation of hypertrophy and tissue integration.

Keywords: Bone marrow stromal cells, Mesenchymal stem cells, Cartilage, Osteoarthritis, Chondrogenesis,Differentiation, Hypertrophy, Oxygen, Micro-pellet

IntroductionDespite considerable investment into bone marrow-derived stromal cells (BMSC, sometimes referred to as“mesenchymal stem cells”) as a cell source for cartilagedefect repair, thus far, no BMSC-based therapies havesuccessfully passed the regulatory and efficacy hurdlesrequired for clinical approval [1]. A major efficacy limi-tation for BMSC-derived engineered cartilage is the ob-served hypertrophic conversion of these cells whentransplanted in vivo in mice, resulting in remodelled,mineralised bone-like tissue [2, 3].The majority of pre-clinical experimentation is per-

formed in small animal models such as mice, rats, or rab-bits [4, 5]. While some orthopaedic studies are performedusing the joints of these small animals as models, thesejoints are anatomically different and experience reducedmechanical forces relative to human joints. Additionally,the thin cartilage in small animal models is capable of re-generation, which differs from cartilage in human joints[4]. In some studies that use immunocompromised smallanimals, human tissue is implanted at ectopic sites, typic-ally in subcutaneous pouches on the backs of mice [2, 3].While this allows for the transplantation of human cells,the implant sites are dissimilar to human joints, as they donot provide mechanical load, and this site is more vascularthan joint capsules. Large animal models such as sheep,pigs, goats, or horses are more appropriate for orthopaedicstudies due to their joint anatomy, and healing propertiesbeing more similar to that of humans [6]. However, unlikeimmunocompromised small animal models, these im-mune competent large animal models require thatspecies-matched cells be used. Therefore, large animalmodels need to be characterised independently to deter-mine their suitability as models for cellular therapies.While sheep are the most commonly used large animalfor the study of cartilage repair [6], the biology of sheep orovine BMSC (oBMSC) remains poorly characterised rela-tive to human BMSC (hBMSC).Our team has developed micro-pellet models to char-

acterise and optimise in vitro hBMSC chondrogenesis [3,7, 8]. Pellet models for studying BMSC chondrogenesiswere first described by Johnstone et al. in 1998 [9].While the traditional pellet model mimics aspects of

developmental mesenchymal condensation, these pellets,formed from ~ 2 × 105 cells each, have a large diameterand suffer steep diffusion gradients and radial tissue het-erogeneity [8]. By contrast, micro-pellets can be formedfrom fewer cells (5 × 103 each), resulting in reduced dif-fusion gradients and more homogeneous tissue. Here-after, traditional pellet cultures assembled from severalhundred thousand cells will be referred to as macro-pellets, while smaller pellets will be referred to as micro-pellets. Previous data demonstrate that micro-pelletsenable characterisation and optimisation of cell culturevariables, such as optimal oxygen concentration [3] orgrowth factor exposure time [7], that are obfuscated bythe radial heterogeneity suffered by traditional macro-pellet models. In addition to being an excellent tissueculture model, micro-pellets could theoretically be usedas building blocks to fill cartilage defects and facilitaterepair [3, 10]. To expedite high-throughput micro-pelletmanufacture, we developed a microwell culture platformcalled the Microwell-mesh [3]. The Microwell-mesh hasa nylon mesh bonded over an array of microwells. Themesh openings are large enough to allow a single cellsuspension to be centrifuged through the mesh, and intothe microwells. Within the first 3 h of culture, cells self-assemble into micro-pellets, becoming too large to es-cape back through the mesh; thus, they are retained indiscrete microwells during weeks of culture manipula-tion. This feature of the Microwell-mesh allows hun-dreds to thousands of micro-pellets to be efficientlycultured simultaneously, facilitating large-scale experi-mentation with micro-pellets that would otherwise betedious or costly.In this study, we aimed to optimise the chondrogenic

culture conditions for oBMSC using the micro-pelletmodel and to evaluate the use of micro-pellets as build-ing blocks to repair osteochondral defects in sheep. Weused the Microwell-mesh platform to manufacturemicro-pellets formed from 5000 oBMSC each and com-pared them to traditional macro-pellets manufacturedfrom 200,000 oBMSC each. Chondrogenic induction wasperformed in atmospheres containing 2%, 5%, or 20%O2, to better characterise this important chondrogenicfactor [8, 11–13]. The resulting cartilage-like tissue was

Futrega et al. Stem Cell Research & Therapy (2021) 12:26 Page 2 of 19

evaluated based on glycosaminoglycan (GAG) accumula-tion, histological analysis, and gene expression. A pilotstudy was then performed to evaluate the capacity ofautologous micro-pellets, manufactured from eitheroBMSC or expanded ovine articular chondrocytes(oACh), to repair osteochondral defects formed in thestifle joint of adult Merino sheep. Two sheep receivedoBMSC micro-pellets, while another two sheep receivedoACh micro-pellets. Each sheep had three replicate de-fects, 6 mm in diameter and 1.5 mm deep. Engineeredconstructs were assembled during the surgery using amould to cast a 6-mm diameter by 1.5 mm deep layer ofmicro-pellets, set with Baxter Tisseel fibrin glue, on aCelGro™ collagen scaffold, and fixed in place with su-tures. The pilot study was terminated 8 weeks later, anddefect infill was characterised via histology.

Materials and methodsLive animal anaesthesia and non-surgical proceduresAll procedures were approved by the Queensland Uni-versity of Technology (QUT) Animal Ethics Committee(approval number 1500001091). For this study, a total ofseven skeletally mature sheep were used. The first threesheep (one accessed through tissue sharing, and two as-sociated with this ethics approval) were used to collectbone marrow aspirates for in vitro studies. For the cartil-age repair studies using autologous cells, four malesheep were used, two were assigned to the oBMSCmicro-pellet group and two were assigned to the oAChmicro-pellet group. Surgical procedures were performedunder general anaesthesia and aseptic conditions andusing a bimodal analgesics regime. For surgical kneeprocedures, analgesia was provided just prior to surgerywith slow release transdermal fentanyl patches (2–3mg/kg/h). A first patch was applied just prior to surgery andtwo more patches were applied in 3-day intervals post-surgery. Trisoprim (1 mL/30 kg/day) was administeredpostoperatively for 2 days to prevent bacterial infection.For euthanasia, sheep were injected in the external jugu-lar with 100 mg/kg Pentobarbital sodium. These proce-dures were performed by a veterinary surgeon and teamat the QUT Medical Engineering Research Facility inBrisbane, Australia. Detailed methods are available uponrequest from the senior corresponding author.

Bone marrow aspiration and BMSC cultureFive different sheep were used for oBMSC in this study.In preliminary optimisation, donors oBMSC 1, oBMSC 2and oBMSC 3 were used. In the cartilage defect repairstudies, oBMSC 4 and oBMSC 5 were used. Bone mar-row aspiration from the iliac crest of sheep was per-formed according to previously described and well-established methods [14]. Briefly, an 11-guage Jamshidineedle was used to aspirate ~ 30mL of bone marrow

from the iliac crest into a 35-mL syringe containing 5mL of heparin (1000 IU/mL). Heparinised bone marrowaspirates were diluted in cell culture medium at a ratioof 5 mL aspirate to 25 mL medium. Growth mediumconsisted of low glucose Dulbecco’s modified Eagle’smedium (LG-DMEM; Thermo Fisher Scientific) supple-mented with 10% fetal bovine serum (FBS; ThermoFisher Scientific), 10 ng/mL fibroblast growth factor-1(FGF-1; Peprotech), 5 μg/mL porcine heparin sodiumsalt (Sigma-Aldrich), and 100 U/mL penicillin/strepto-mycin (PenStrep; Thermo Fisher Scientific). Cells wereseeded into Nunc™ T175 cm2 tissue culture flasks(Thermo Fisher Scientific) and incubated overnight in anormoxic incubator (20% O2) with 5% CO2 at 37 °C. Fol-lowing overnight incubation, media was exchanged toremove non-adherent cells. Cultures were moved to ahypoxic incubator (2% O2, 5% CO2) following the 24-hattachment period. Adherent cells were cultured until80% confluency, enzymatically harvested (0.25% Trypsin/EDTA, Thermo Fisher Scientific), and reseeded at 1500cells/cm2 in new T175 flasks. Culture medium was ex-changed twice weekly. Cells were cryopreserved at lowpassage (P0-P2) in 90% FBS and 10% DMSO (Sigma-Aldrich).hBMSC isolation and expansion was performed as pre-

viously described [5]. Bone marrow aspirates were col-lected from the iliac crest of fully informed andconsenting healthy human volunteer donors. The MaterHealth Services Human Research Ethics Committee andthe Queensland University of Technology Human EthicsCommittee approved aspirate collection (Ethics number:1541A). hBMSC were cultured in the same way asoBMSC.

Flow cytometry characterisationFor flow cytometry analysis, oBMSC were stained withthe following anti-human antibodies: CD34-APC, CD90-FITC, CD73-APC, CD105-PE, CD146-APC, CD271-PE,HLA-DR-PE, and isotype controls (Miltenyi Biotec).Additionally, the following anti-sheep antibodies wereused: CD45-FITC and CD44-FITC (Bio-Rad). Cells werestained as per the manufacturers’ instructions, and ana-lysis was completed using an LSR II flow cytometer (BDBiosciences). In this analysis, we also included sheepbone marrow mononuclear cells (MNC) and hBMSC.These cell populations served as controls for antibodyperformance against a common hBMSC population, andfor sheep MNC, where haematopoietic cells (CD45+)would be expected. Data were analysed using FlowJosoftware, version 10 (BD Biosciences).

Articular cartilage biopsy and oACh cultureTwo oACh donors (oACh 1 and oACh 2) were used inthe cartilage defect repair component of this study.


Articular cartilage biopsies were harvested from the fem-oral trochlear grooves. The knee joint capsule wasaccessed via a medial parapatellar incision (~ 5 cm). Thepatella was displaced laterally and retained in a proximalposition with small wound retractors. The joint capsulewas opened to expose the femoral trochlear groove. Fivethin (~ 1mm thick and 4–6mm wide) shavings of ar-ticular cartilage were aseptically removed with a scalpeland transferred to a sterile specimen jar containing LG-DMEM with PenStrep and 1X Antibiotic-Antimycotic(Anti-Anti; Thermo Fisher Scientific). The biopsied tis-sues were placed in a Petri dish and chopped upcoarsely. The tissue was resuspended in 10 mL of diges-tion solution containing type II clostridial collagenase(~ 430 U/mL; Sigma-Aldrich), PenStrep, and Antibiotic-Antimicotic (Gibco) and incubated overnight in anormoxic incubator (20% O2) with 5% CO2 at 37 °C.Following digestion, the cells were washed twice bycentrifugation (500×g) with LG-DMEM containing 10%FBS. oACh were grown in a hypoxic (2% O2) incubatorin growth medium as described above for oBMSC.

Fabrication of the Microwell-mesh culture deviceThe Microwell-mesh was prepared as detailed in a previ-ous paper published by our laboratory [3]. Briefly, a ~ 4-mm layer of polydimethylsiloxane (PDMS; Dow Corn-ing) was poured into a polystyrene negative templatethat had an inverted microwell pattern. PDMS was curedat 80 °C for 60 min. A wad punch was used to creatediscs that fit snugly into Nunc™ 6-well plates (ThermoFisher Scientific). Individual microwells measured 2mm × 2mm with a depth of 0.8 mm [3]. A nylon mesh(36-μm square pore openings, part number CMN-0035,Amazon.com) was bonded to the open face of the PDMSdiscs with silicone glue (Selleys, Aquarium Safe). Plateswere sterilised by submerging in a 70% ethanol solutionfor a minimum of 30min and rinsed 3 times withphosphate-buffered saline (PBS; Thermo Fisher Scien-tific). Prior to cell seeding, microwells were rinsed with asterile 5% Pluronic F-127 solution (Sigma-Aldrich) inPBS to render the PDMS surface non-adhesive and topromote cell aggregation [15, 16]. Wells were rinsed 3times with PBS to remove excess Pluronic, and then,cells were seeded as described below.

Osteogenic and adipogenic inductionTo induce osteogenesis and adipogenesis, oBMSC wereseeded at 30,000 cells/cm2 in standard tissue culture wellplates and cultured for 21 days in the respective induc-tion media. Osteogenic induction medium wascomposed of 10 mM β-glycerol phosphate, 100 nM dexa-methasone (Dex), 50 μM L-ascorbic acid 2-phosphate (allfrom Sigma-Aldrich), 10% FBS, and PenStrep in highglucose (HG)-DMEM (Thermo Fisher Scientific).

Conventional adipogenic induction medium was com-posed of 10 μg/mL insulin, 100 nM Dex, 25 mM indo-methacin, and 3-Isobutyl-1-methylxanthine (IBMX) (allfrom Sigma-Aldrich), 10% FBS, and PenStrep in HG-DMEM. We noticed that adipogenic induction was notsuccessful using conventional adipogenic inductionmedium, which appeared to be toxic. We then testedwhether a single ingredient could be removed from theadipogenic induction medium to eliminate the toxicity,and if this modified medium would support oBMSC adi-pogenic induction. We included a hBMSC population inthis adipogenic assay for comparison.To assess osteogenic and adipogenic induction, mono-

layers were fixed for 20 min in 4% paraformaldehyde(PFA), then washed and stained. Mineralised matrix de-position was assessed in osteogenic cultures usingAlizarin Red S staining (Sigma-Aldrich). Lipid vacuolesin adipogenic cultures were stained using Oil Red O(Sigma-Aldrich). Wells were rinsed with distilled water,incubated with Alizarin Red S or Oil Red O stain for 10min, then washed with water and visualised.

Chondrogenic induction culturesTo induce chondrogenesis, cells were resuspended inchondrogenic medium composed of HG-DMEM, Pen-Strep, GlutaMax, 1X ITS-X, 100 μM sodium pyruvate(all from Thermo Fisher Scientific), 10 ng/mL TGF-β1(PeproTech), 100 nM Dex, 200 μM ascorbic acid 2-phosphate, and 40 μg/mL L-proline (Sigma-Aldrich).Prior to cell seeding, 3 mL of cell-free chondrogenic in-duction medium was added to each well, and plates werecentrifuged for 5 min at 2000×g to ensure elimination ofair bubbles from microwells in the Microwell-mesh. Togenerate micro-pellets, each well was seeded with 1.2 ×106 oBMSC in 1mL of chondrogenic medium and theplate was centrifuged at 500×g for 3 min to force thecells through the nylon mesh and into the microwells.The microwell array had ~ 250 microwells per well, andtherefore, ~ 5000 cells were seeded per micro-pellet. Atculture harvest, the nylon mesh was peeled off of thePDMS microwell insert to liberate micro-pellets. Controlmacro-pellet cultures were established by seeding 2 ×105 oBMSC in 1mL of induction medium in 96-well,deep V-bottom plates (Corning). Cultures were main-tained at 2%, 5%, or 20% O2, and 5% CO2 in a 37 °C in-cubator for optimisation experiments, where indicated.Media was exchanged every second day.

Quantification of glycosaminoglycans (GAG) and DNATissues were incubated in an overnight papain digest(1.6 U/mL papain, 10 mM L-cysteine; both from Sigma-Aldrich) in a 60 °C water bath. The 1,9-dimethylmethy-lene blue (DMMB, Sigma-Aldrich) assay was used toquantify GAG in the digested tissues [8]. Chondroitin


http://amazon.com

sulfate from shark cartilage (Sigma-Aldrich) was used togenerate a standard curve. DNA content in the tissueswas quantified using the Quant-iT™ PicoGreen® dsDNAassay kit (Thermo Fisher Scientific) as per the manufac-turer’s protocol.

Quantitative PCR (qPCR)At harvest, tissues were collected and stored in 350 μLRLT buffer (Qiagen) containing β-mercaptoethanol(Sigma-Aldrich) at − 80 °C. The samples were crushed in1.5-mL microcentrifuge tubes using sterile micropestles;then, RNA was isolated using the RNeasy Mini Kit(Qiagen) with on-column DNase I (Qiagen) digestion, asper the manufacturer’s instructions. RNA was quantifiedwith a NanoDrop Lite spectrophotometer (ThermoFisher Scientific), and RNA was reverse-transcribedusing the SuperScript III First-Strand Synthesis Systemfor RT-PCR (Thermo Fisher Scientific) as described bythe manufacturer. The master mix included 2X SYBRGreen PCR Master Mix (Applied Biosystems), 200 nM ofthe forward and reverse primers, RNase-free water, and1 μl of sample cDNA. The 5 μl reactions were run intriplicate in a 384-well plate inside a Viia7 Real TimePCR System (Applied Biosystems). The initial cycle was50 °C for 2 min and 95 °C for 10 min, followed by 40 cy-cles of 95 °C for 15 s and 60 °C for 1 min. The melt curveand electrophoretic gels were evaluated to confirm thespecificity of products. Primer set information for genesof interest is provided in Table 1. Gene expressionvalues were normalised to GAPDH and calculated usingthe ΔCt method.

Histology and immunohistochemistryTissues were washed in PBS and fixed in 4% PFA for 20min and frozen in Tissue-Tek OCT compound (SakuraFinetek). Samples were cryosectioned at 7 μm and col-lected onto poly-lysine coated slides and then frozenuntil further processing. The sectioned tissues were fixedfor 15 min with 4% PFA and washed with PBS. Sectionswere stained with Alcian blue (Sigma-Aldrich) for 30min to visualise GAG distribution. Slides were rinsedwith tap water and sections were counterstained with

Nuclear Fast Red for 5 min, rinsed with tap water, andmounted and coverslipped. For immunohistochemicalstaining, sections were treated with hyaluronidase (2 U/mL, Sigma-Aldrich) for 30 min at 37 °C. Slides werewashed twice with 0.025% Triton X-100/PBS, thenblocked for 60 min at room temperature with 10% nor-mal goat serum (ThermoFisher Scientific). Primary anti-bodies (all from Abcam) raised against collagen type I(1:800; ab6308), type II (1:100; ab34712), and type X (1:100; ab58632) were diluted in 1% BSA/PBS and incu-bated on sections at 4 °C, overnight. Sections were rinsedtwice for 5 min with 0.025% Triton X-100/PBS. Slideswere incubated with 0.3% H2O2 for 15 min and rinsedtwice with PBS. Sections were incubated for 60 min atroom temperature with secondary antibodies, goat anti-rabbit (HRP; ab6721) or goat anti-mouse (HRP;ab97023), at 1:1000 (Abcam) in 1% bovine serum albu-min (BSA)/PBS, then washed twice with PBS. A DABchromogen kit (Abcam) was used to develop the signalfor 8 min; then, slides were rinsed in water, mounted,and coverslipped.

Autologous repair of in vivo cartilage defects in sheepWe evaluated oBMSC micro-pellets and oACh micro-pellets in vivo, in sheep cartilage defects. Due to the highcost of sheep and the complexity of the procedure, weperformed a pilot study with 2 sheep for oBMSC and 2sheep for oACh micro-pellet evaluation. The procedurewas carried out in two steps. In a first procedure, bonemarrow aspirates or articular cartilage were collectedfrom 2 sheep each (total of 4 sheep). The cells wereisolated, expanded, and frozen at P0, as described above.In a second procedure, cells were thawed, expanded forone additional passage, induced for 10 days inMicrowell-mesh cultures, and implanted in cartilage de-fects created in the trochlear groove of knees of thesame sheep from which the original cells were sourced(autologous implantation). All cell culture was carriedout in an incubator set at 2% O2, 5% CO2, and 37 °C.Sheep that received oACh micro-pellets received the im-planted tissue in the knee opposite to the one where theoriginal cartilage biopsy had been collected. Some

Table 1 Primers used for qPCR for ovine genes

Gene Forward Reverse Amplicon (bp) Accession #

ACAN TTTGGACTTTGGCAGAATACC AATCCAGAAGGAAGACCACTTG 78 FJ200438.1

COL2A1 CTGTCCTTCGGTGTCACGG CGGGCTTCCACACATCCTTAT 93 XM_027967399.1

COL1A1 CAGGGAGACAGAGGCATCAAG ATCTTTGCCAGGAGTACCAGC 156 XM_027974706.1

COL10A1 GCATAAAAGGCCCACCATCC CTGGTGGACCGGGGATAC 88 XM_004011185.4

SOX9 CAAGCTCTGGAGACTGCTGAA CCGTTCTTCACCGACTTCCT 135 XM_027974011.1

RUNX2 CGCCTCACAAACAACCACAG GCTTGCAGCCTTAAATGACTCT 143 XM_027959124.1

GAPDH ACAGTCAAGGCAGAGAACGG CCAGCATCACCCCACTTGAT 98 NM_001190390.1


micro-pellets were saved for GAG staining and qPCRanalysis. Sheep were anaesthetised and prepared for sur-gery. Following exposure of the femoral trochleargroove, three 6-mm defects were made in the trochleargroove using a 6-mm biopsy punch, and the articularand calcified cartilage was removed using a curette untilthe subchondral bone was exposed (~ 1.5 mm deep). Ap-proximately 5 mm of intact cartilage was left betweendefects. The defect sites and exposed cartilage were irri-gated with saline regularly to clean and maintain tissuemoisture.To prepare the cartilage constructs, we used custom

well chambers that consisted of a glass slide and a cy-lindric well made from a PDMS ring that had an in-ternal diameter of 6 mm and 1.5 mm depth. The wellchambers were sterilised by autoclaving and handledaseptically. A 6-mm wide disc was punched from asheet of CelGro™ (Orthocell Ltd., Perth, Australia) col-lagen scaffold and placed in the bottom of the well. Adropper was used to drop micro-pellets into the wellon top of the CelGro scaffold. Excess liquid was re-moved from the chamber with a pipette tip. Once thewell was adequately filled with micro-pellets, twodrops of fibrinogen were applied, followed by twodrops of thrombin, both from a Baxter Tisseel fibringlue kit. The gelled construct was removed from thechamber using forceps and inserted into the cartilagedefect with the micro-pellets facing the subchondralbone and the CelGro scaffold sitting on top, adjacentto the superficial cartilage layer. To provide a scaffold-only control, one defect site in both the oBMSC andoACh sheep groups was filled with fibrin glue and theCelgro scaffold, but no cells. Four degradable sutureswere used to fix the membrane to adjacent cartilagetissue. The patella was returned to its natural positionand the incision was sutured to close the knee joint.Knee joints were treated with antibiotic and bandagedfor the oACh animals, and a plaster cast was addition-ally applied on the oBMSC animals. Plaster casts wereintroduced in the oBMSC group, in an effort to fur-ther reduce leaning on the operated joint, which wasobserved with the first group of animals (oAChgroup). Both groups of sheep recovered in slings for 2weeks in an effort to limit the weight on the repairedjoint and then were permitted to roam freely for theremaining 6 weeks. A total of 8 weeks after the proced-ure, sheep were euthanised and joints recovered. Thedefect regions were trimmed with a saw and fixed in4% PFA for 72 h. The defects were decalcified using aKOS rapid decalcification machine (Milestone). Thetissues were dehydrated, paraffin embedded, and sec-tioned at 5 μm. The tissues were stained with haema-toxylin and eosin (H&E), as well as Alcian blue andNuclear Fast Red.

Statistical analysisStatistical significance of quantitative data was per-formed using ANOVA in GraphPad Prism. Statisticalsignificance was set at P < 0.05. For GAG, DNA, andqPCR analysis, each culture was performed for 4 repli-cate wells (n = 4). The number of unique cell donors oranimals was as indicated in the “Materials and methods”and “Results” sections.

ResultsoBMSC characterisationIsolated oBMSC appeared spindle-shaped upon expansion(Fig. 1b), resembling BMSC from other species such ashBMSC. Flow cytometry analysis is summarised in Fig. 1afor oBMSC populations derived from three unique donors,one sheep bone marrow mononuclear cell (MNC) donorpopulation, and one hBMSC donor population (histogramsare shown in Supplementary Figures 1, 2, 3, 4 and 5). TheFig. 1a upper graph displays the percentage of cells thatexpressed the antigen, while the lower graph displays therelative mean fluorescence intensity (MFI) of each antibodyon each cell population. Unlike hBMSC, oBMSC containedonly a small population of cells positive for CD73 (10.2% ±12.7%), were negative for CD90 and CD105, and containeda small population positive for CD34 (13.1% ± 3.7%). LikehBMSC, oBMSC were 100% positive for CD44, mostlypositive for CD146 (69% ± 8.4%), and had a small popula-tion that was positive for CD271 (17.0% ± 4.3%). The lowfrequency of CD45+ cells (3.1% ± 1.1%) in oBMSC cultures,along with data indicating that 76% of sheep MNCs stainedpositively for CD45, suggests that the CD45 antibodyworks, and that the expanded oBMSC had been depletedof CD45+ cells. Given the successful depletion of CD45+

cells during oBMSC enrichment and expansion, the smallpopulation of CD34+ cells (13.1% ± 3.7%) likely reflectsnon-specific binding with this antibody. While the anti-body panel had limited cross-reactivity with oBMSC, therelative enrichment of CD146+ cells [17] and depletion ofCD45+ cells [18] suggested that these cells were suitablefor use in further characterisation studies.The multipotency of oBMSC was evaluated using in vi-

tro tri-lineage differentiation assays, following protocolsconventionally used for hBMSC. Alcian blue staining ofoBMSC macro-pellets cultured in chondrogenic induc-tion medium demonstrated deposition of cartilage-likeextracellular matrix (ECM; Fig. 1c). Alizarin Red S stain-ing confirmed calcium nodules in osteogenically inducedcultures (Fig. 1d). Following adipogenic induction usingconventional induction medium, we noticed that adipo-genesis was poor and the cells appeared unhealthy anddetached from the well plates. To test whether there wasa component in the conventional adipogenic mediumthat was toxic to oBMSC, we set up an experiment inwhich we removed one component at a time from the


conventional adipogenic medium. We also included anhBMSC population in this test as a control. We foundthat removal of IBMX resulted in healthy oBMSC cul-tures that underwent adipogenesis, forming ample lipidvacuoles, demonstrated by Oil Red O staining (Fig. 1e).Removal of IBMX from hBMSC adipogenic inductioncultures did not demonstrate the same stark difference,relative to conventional adipogenic induction medium.

Growth of macro-pellets and micro-pellets in themicrowell-mesh systemTo characterise the chondrogenic potential of oBMSC,we assessed chondrogenesis in both our customMicrowell-mesh culture platform (5 × 103 cells/micro-pellet, Fig. 2b), as well as traditional macro-pellet culture(2 × 105 cells/macro-pellet, Fig. 2a). Over 14 days ofchondrogenic induction culture, oBMSC micro-pellets

Fig. 1 oBMSC characterisation. a Flow cytometry analysis of three unique oBMSC donors (black circles), with ovine MNC (oMNC, red circles) andhBMSC (blue circles) as control populations. The upper graph shows the percentage of cells that were positive for the markers. The lower graphshows the mean fluorescence intensity (MFI) of the markers for each cell population. Flow cytometry histograms for each donor and each markerare shown in Supplementary Figures 1, 2, 3, 4 and 5. b Bright-field image of oBMSC during expansion. c Chondrogenic induction shown byAlcian blue staining of matrix GAG. d Osteogenic induction of oBMSC showing mineralised nodules stained with Alizarin Red S. e Adipogenicinduction of three unique oBMSC donors and an hBMSC donor, showing that conventional induction media does not result in adipogenesis ofoBMSC unless IBMX is removed. Lipid vacuoles are stained red with Oil Red O. Scale bar = 400 μm in b and c, 2 mm in d, and 100 μm in e


and macro-pellets in 2%, 5%, and 20% O2 atmospheresshowed an increase in size (Fig. 2c). Micro-pellets cul-tured at 2% O2 were larger than micro-pellets culturedat 5% O2 or 20% O2, and this was consistent across 3unique oBMSC donors (see Supplementary Figure 6 fortwo additional donors). Some variability was observedacross different oBMSC donors, particularly in macro-pellet cultures (compare Fig. 3d with Supplementary Fig-ure 6A and C). However, this is not surprising as donorvariability has been well-documented with hBMSC [19].

Low oxygen increases GAG production in micro-pelletsGAG production increased from day 7 to day 14 forboth macro-pellets and micro-pellets, for both of thetwo unique oBMSC donors analysed, oBMSC 1 and

oBMSC 2 (Fig. 3a and b). In macro-pellet cultures, theamount of GAG and DNA trended toward a higher levelat lower O2 concentrations (2% and 5%), but this wasnot consistently statistically significant due to a largeamount of variability between individual macro-pellets(Fig. 3a, c, and e). In micro-pellet cultures, the amountof GAG was greater at lower O2 concentrations (2% and5%) and this was statistically significant for oBMSC 1 onday 7 and day 14, and for oBMSC 2 on day 14 (Fig. 3b).The amount of DNA in micro-pellets was higher foroBMSC 1 at lower O2 concentrations (2% and 5%), com-pared with 20% O2, but the magnitude of differences be-tween O2 concentrations was generally small (Fig. 3d).When GAG was normalised to DNA for micro-pelletcultures, the amount of GAG/DNA was greater at lower

Fig. 2 Schematic of macro-pellet and micro-pellet assembly. a Macro-pellets were assembled by centrifuging 2 × 105 oBMSC in V-bottom deepwell plates. b Micro-pellets were assembled by centrifuging 5 × 103 oBMSC per microwell, in the Microwell-mesh (a and b adapted from [7]).Microscopic images over a 14-day culture period of c macro-pellets cultured in deep-well plates and d micro-pellets cultured in the Microwell-mesh platform. Scale bar = 1 mm. Images for 2 additional oBMSC donors are shown in Supplementary Figure 6


O2 concentrations (2% and 5%), compared with 20% O2,and this was statistically significant for both oBMSC do-nors at both time points (Fig. 3f). There was substantialvariability observed between the two oBMSC donors,with oBMSC 1 producing substantially more GAG thanoBMSC 2 when assessed at both day 7 and day 14.As previously observed for hBMSC [3, 7], GAG/DNA

production in micro-pellet cultures was typically greaterthan in macro-pellet cultures. At 5% O2, GAG/DNA was1.44 ± 0.11-fold greater (P = 0.0039) in oBMSC 1 micro-pellet cultures than macro-pellet cultures on day 14, andfor oBMSC 2, GAG/DNA was 2.27 ± 0.42-fold (P <0.0001) greater in micro-pellet cultures than macro-pellet cultures on day 7. At 2% O2, GAG/DNA wasgreater in micro-pellet cultures than in macro-pellet cul-tures for both donors, and on both days 7 and 14. ForoBMSC 1 at 2% O2, GAG/DNA was 2.68 ± 0.34-fold

(P < 0.0001) and 1.60 ± 0.06-fold (P < 0.0001) greater inmicro-pellet cultures compared with macro-pellet cul-tures on days 7 and 14, respectively. For oBMSC 2 at 2%O2, GAG/DNA was 4.90 ± 0.96-fold (P < 0.0001) greateron day 7 and 3.22 ± 0.63-fold (P < 0.0001) greater on day14 in micro-pellet cultures compared with macro-pelletcultures.

ECM characterisation in macro-pellets and micro-pelletsWe assessed the production of cartilage-like ECM with 3unique oBMSC donors in macro-pellets and micro-pellets cultured in 2%, 5%, and 20% O2 atmospheres, onday 7 and day 14. Alcian blue staining revealed increasedGAG matrix accumulation in both macro-pellets andmicro-pellets cultured at 2% O2 compared with 20% O2(Fig. 4a and b). On day 7, macro-pellets contained cell-dense cores (red stained nuclei) with little cartilage-like

Fig. 3 Quantification of GAG and DNA in macro-pellets and micro-pellets on day 7 and day 14 for oBMSC 1 and oBMSC 2. Quantities of GAG in amacro-pellets and b micro-pellets. Quantities of DNA in c macro-pellets and d micro-pellets. GAG normalised to DNA in e macro-pellets and fmicro-pellets (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001)


matrix at all oxygen levels (Fig. 4a). For this reason, theNuclear Fast Red staining is prominent at these earlytime points. By day 14, macro-pellets exhibited moreuniform GAG matrix throughout their diameter foroBMSC 1 at all oxygen concentrations, and for BMSC 2and BMSC 3 at 2% O2. Similarly, micro-pellets culturedat 2% O2 stained more uniformly with Alcian bluethroughout their diameter by day 14, compared with 5%and 20% O2 (Fig. 4b). Consistent with GAG quantifica-tion (Fig. 3), variability was also evident in Alcian blueGAG staining between unique oBMSC donors.We also performed immunohistochemistry staining

for cartilage-associated ECM molecule, type II collagen(Supplementary Figure 7); bone-associated molecule,type I collagen (Supplementary Figure 8); andhypertrophy-associated molecule, type X collagen (Sup-plementary Figure 9). While we observed some stainingfor all of these collagen molecules, we did not observe astaining pattern that correlated with differing oxygenconcentrations.

In vivo cartilage defects in sheepWe evaluated the potential of oBMSC micro-pellets torepair cartilage defects in sheep. Micro-pellets inducedfor 10 days were suspended in fibrin glue and anchoredto a collagen Celgro scaffold/membrane that was used tokeep the tissues fixed in osteochondral defects in sheep

(see schematic in Fig. 5). Three defects (6 mm in diam-eter and 1.5 mm deep) were made in the trochleargroove of one hind leg for each sheep and filled with re-pair tissues. Two sheep were implanted with micro-pellets made from induced oBMSC, and two sheep wereimplanted with micro-pellets made from induced oACh,as controls. Sheep were euthanised after a total of 8weeks from the day of receiving an implant, and defectrepair was assessed by histology.

Histological and qPCR assessment of oBMSC and oAChmicro-pellets pre-implantWe assessed the micro-pellets implanted in sheep histo-logically and for relative gene expression (Fig. 6). Ten-day, chondrogenically induced micro-pellets derivedfrom both oBMSC (Fig. 6a) and oACh (Fig. 6b) donorsstained for GAG with Alcian blue, demonstrating the de-velopment of cartilage-like tissue. Histologically, theoBMSC- and oACh-derived micro-pellets were visuallyindistinguishable from each other. qPCR analysis wasperformed to quantify the relative gene expression (Fig.6c) for cartilage-associated genes (COL2A1, ACAN, andSOX9) and for bone or hypertrophy-associated genes(COL1A1, COL10A1, and RUNX2) in both oBMSC- andoACh-derived micro-pellets, at day 0, before induction,and on day 10 of induction. For both oBMSC- andoACh-derived micro-pellets, the expression of COL2A1,

Fig. 4 Alcian blue staining for GAG in a macro-pellet and b micro-pellet sections. Nuclei are stained red. Scale bar = 400 μm


ACAN, and SOX9 were statistically significantly in-creased on day 10 following chondrogenic induction,relative to day 0. For oBMSC-derived tissues, COL2A1,ACAN, and SOX9 increased by 1.45 × 106 ± 3.82 × 105-fold, 2.56 × 104 ± 7.38 × 103-fold, and 61.3 ± 23.6-fold, re-spectively, on day 10 relative to day 0. For oACh-derivedtissues, COL2A1, ACAN, and SOX9 increased by 53.8 ±10.7-fold, 13.5 ± 5.11-fold, and 4.23 ± 2.29-fold, respect-ively, on day 10 relative to day 0. ACAN expression was1.65 ± 0.624 higher in oACh micro-pellets following in-duction (day 10) relative to oBMSC micro-pellets (day10), whereas COL2A1 and SOX9 were statistically similarbetween oBMSC and oACh at both time points. ForoBMSC-derived tissues, the expression of COL1A1(21.3 ± 3.41-fold) and COL10A1 (1.61 × 103 ± 283-fold)were increased following chondrogenic induction, whileRUNX2 expression was not changed following induction.Following induction (day 10), COL1A1 (8.97 ± 1.44-fold)

and COL10A1 (1.10 × 103 ± 193-fold) were higher inoBMSC-derived micro-pellets relative to oACh-derivedmicro-pellets. RUNX2 expression was higher in oBMSC-derived micro-pellets compared with oACh micro-pellets, both on day 0 (345 ± 213-fold) and on day 10(708 ± 254-fold).

Histological assessment of oBMSC and oACh micro-pelletspost-implantFollowing 8 weeks of incubation in vivo in sheep, jointswere harvested, and defect repair was characterised his-tologically (Figs. 7 and 8). None of the oBMSC or oAChmicro-pellet groups showed complete cartilage fill orseamless integration with adjacent tissue. Figure 7 showsthree representative defects that were filled withoBMSC-derived micro-pellets and stained with H&E andAlcian blue. H&E staining showed that most of the de-fects were filled with fibrous tissue, vascular tissue, and

Fig. 5 Schematic of in vivo evaluation of micro-pellets. a Ten-day chondrogenically induced oBMSC or oACh micro-pellets were layered onto aCelgro scaffold in a custom-made cylindric chamber (6 mm in diameter and 1.5 mm high). b Two drops of fibrinogen were added to the micro-pellets, followed by two drops of thrombin, to adhere the micro-pellets to each other and to the Celgro scaffold. c The repair construct was liftedwith forceps. d The repair construct was placed into a cartilage defect that was of a similar dimension, with the micro-pellet layer facing downinto the subchondral layer and the Celgro scaffold on top. e Four sutures were used to fix the Celgro scaffold to the adjacent cartilage tissue. fSheep were placed in slings for 2 weeks to recover and then were allowed to roam freely for the remaining 6 weeks


blood cells (Fig. 7a–c). There was faint Alcian blue stain-ing for GAG in some of the defects (see Fig. 7b, lowerleft panel), suggesting either residual cartilage from theimplanted micro-pellets, or that cartilage repair was oc-curring very slowly. This could also have been from infil-tration of chondroprogenitors from adjacent tissuerather than oBMSC. Supplementary Figure 10 showshistology of a scaffold-only control, where the defect wasfilled with fibrin glue and the Celgro scaffold, but nocells. Of the 5 defects filled with oBMSC micro-pellets,we could only find one defect that still had some recog-nisable micro-pellets, which are shown Fig. 7c. Based onthe morphology of these oBMSC micro-pellets, it ap-peared as though they may have been in the process ofdedifferentiating or being absorbed by surroundingtissue.Figure 8 shows three representative defects that were

filled with oACh-derived micro-pellets and stained with

H&E and Alcian blue. H&E staining showed that mostof the defect was filled with fibrous tissue, but unlike de-fects filled with oBMSC micro-pellets, they generally ap-peared to have attracted fewer blood lineage cells andless vascular tissue. Of the 5 defects filled with oAChmicro-pellets, we could only find two defects that stillhad substantial cartilage-like tissue fill, which are shownin Fig. 8b and c. There was still a significant amount offibrous tissue observed in these defects; however, theyshowed the most promising result in terms of theamount of Alcian blue-stained GAG matrix and morph-ology of the cells, which appeared to form lacunae. Sup-plementary Figure 10 shows histology from an emptyScaffold-only control defect site. While oACh-derivedmicro-pellets yielded more promising histology, macro-scopic imaging of the joint revealed that defectsremained visible, with substantial repair visible only in asingle defect (Supplementary Figure 11).

Fig. 6 Characterisation of induced micro-pellets used for sheep studies. a oBMSC and b oACh micro-pellet sections stained with Alcian blue forGAG following 10 days of chondrogenic induction (scale bar = 500 μm and inset scale bar = 100 μm). c qPCR analysis of cartilage-like tissuesformed from oBMSC and oACh after expansion (day 0) and 10 days after chondrogenic induction (day 10). Relative gene expression was assessedfor chondrogenic genes COL2A1, ACAN, and SOX9, and for hypertrophic genes COL1A1, COL10A1, and RUNX2. Gene expression levels werenormalised to GAPDH and represent 2^-ΔCt values. The two unique donors are distinguished by black versus grey symbols. The mean isrepresented by the horizontal line (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001)


DiscussionIn this study, we characterised oBMSC expansion cul-tures in vitro, used a micro-pellet model to identify theoptimal oxygen atmosphere for oBMSC chondrogenesis,and then compared the capacity of micro-pellets assem-bled from either oBMSC or oACh to repair osteochon-dral defects in adult sheep.oBMSC were successfully isolated using plastic adhe-

sion and cells exhibited a spindle-shaped morphologysimilar to hBMSC. Expanded oBMSC were characterisedusing a flow cytometry antibody panel commonly usedto characterise hBMSC [20, 21], including CD73, CD90,CD105, CD44, CD146, CD271, CD45, CD34, and HLA-

DR. Like our control hBMSC population, expandedoBMSC were essentially negative for haematopoieticmarkers (CD45 and CD34), strongly positive (~100%)for CD44, contained a large population (~ 70%) ofCD146+ cells, and had a very small population (< 3%) ofCD271+ cells. CD146 is viewed as one of the definingmarkers of hBMSC and is associated with hBMSC cap-acity to form ectopic marrows and support haematopoi-etic stem cells [22, 23]. CD271 has been reported to beexpressed on the surface of hBMSC [24] and oBMSC[25]. However, CD271+ cell number in hBMSC culturesdeclines rapidly in response to medium serum supple-mentation [26], and our hBMSC control population and

Fig. 7 Representative H&E and Alcian blue stained images of in vivo oBMSC micro-pellet repair tissues after 8 weeks. Repair tissue was largelycomposed of a fibrous tissue and b fibrous tissue with significant vascular and blood cell infiltration (triangle). c In one of 6 defects, a few micro-pellets could be identified; the boxed regions are enlarged to the right, with arrows pointing to micro-pellets. The scaffold was largely degraded,but could be seen in some defects (see asterisks in b and c). Boxes with dashed lines in images on the left are enlarged in the middle (H&E) andright (Alcian blue) images. Scale bars in left panels = 1 mm. Scale bars in middle and right panels = 100 μm


oBMSC cultures also had few CD271+ cells. We did notobserve positive staining of oBMSC using anti-humanCD90 or CD105 antibodies (< 3%), and only two oBMSCdonors stained weakly for CD73 (< 30%), while onedonor was essentially negative for CD73 (< 3%). A previ-ous study also reported poor cross-reactivity of oBMSCsurface antigens and anti-human CD73, CD90, andCD105 antibodies [27]. As expected, oBMSC were nega-tive for the human-specific marker, HLA-DR, whichserved as a negative control. Other markers that havebeen positively detected on oBMSC include CD166 andCD29 [28]. While the full panel of hBMSC antibodymarkers did not react against oBMSC, we were able to

confirm the effective depletion of haematopoietic cells,and that the plastic-enriched cells displayed some over-lap with known hBMSC antigens, which can be used asputative oBMSC markers in future studies.Using conventional tri-lineage induction media,

oBMSC underwent osteogenic and chondrogenic induc-tion successfully, while initially, adipogenic inductionfailed. Using conventional adipogenic induction medium,we observed substantial cell death. By removing onecomponent of the induction medium at a time, we iden-tified that IBMX was toxic to oBMSC. Exclusion ofIBMX from the conventional adipogenic inductionmedium resulted in robust adipogenesis of oBMSC.

Fig. 8 Representative H&E and Alcian blue-stained images of in vivo oACh micro-pellet repair tissues after 8 weeks. Repair tissue was largelycomposed of a fibrous tissue with some vascular tissue (triangle), with b one defect showing modest regions of GAG-rich repair cartilage(arrows), and c another defect showing more substantial regions of GAG-rich repair cartilage (arrows). Boxes with dashed lines in images on theleft are enlarged in the middle (H&E) and right (Alcian blue) images. Scale bars in left panels = 1 mm. Scale bars in middle andright panels = 100 μm


IBMX is a competitive non-selective phosphodiesteraseinhibitor [29], commonly included in BMSC adipogenicinduction medium [30]. While other studies do not spe-cifically report the toxicity of IBMX during oBMSC adi-pogenesis, a previous publication showed an image of asingle cell from hBMSC and oBMSC cultures after 14days of induction in IBMX-supplemented medium [28,31]. In these images, the adipogenic induced hBMSC ap-peared healthy and contained many lipid vacuoles, whilethe oBMSC appeared relatively small with few lipid vac-uoles [31]. Another study induced oBMSC in mediumsupplemented with IBMX and displayed cells in a rela-tively sparse monolayer, with few lipid vacuoles [32],potentially indicating some level of toxicity. These out-comes are similar to our own observations where only afew small oBMSC remained in adipogenic cultures sup-plemented with IBMX. For effective oBMSC adipogenicinduction, exclusion of IBMX from induction culturesappears to be necessary. Why oBMSC respond differ-ently to IBMX during adipogenic induction, comparedwith hBMSC, is not immediately obvious. This toxicity,however, identifies an important difference in the biol-ogy of oBMSC and hBMSC and highlights the import-ance of characterising and comparing cells of differentspecies independently.We characterised the capacity of oBMSC to form

cartilage-like tissue in micro-pellet and traditionalmacro-pellet cultures, in 2%, 5%, or 20% O2 atmo-spheres. A previous study that included both an oBMSCpellet model and mathematical modelling concluded thatoptimal oBMSC chondrogenesis occurs at 10–11% O2[33]. Another study demonstrated that superior chon-drogenesis could be achieved if the oBMSC were ex-panded in a 5% O2 atmosphere, rather than at 20% O2atmosphere [34]. In this previous study, oBMSC ex-panded in 5% or 20% O2 atmospheres were assembledinto macro-pellets for chondrogenic differentiation [34].In our study, we expanded all oBMSC in a 2% O2 atmos-phere and observed incrementally greater GAG produc-tion when oBMSC were assembled into a micro-pelletmodel and cultured at 2% O2, compared with 5% or 20%O2 atmospheres. GAG/DNA output was greatest inmicro-pellets cultured in 2% O2 atmosphere. In macro-pellet cultures, GAG/DNA content trended towardhigher levels in lower O2 atmospheres, but this was notconsistently statistically significant, and greater variabil-ity in data was observed in macro-pellet cultures. At 2%O2, GAG/DNA was significantly greater in micro-pelletcultures relative to macro-pellet cultures. This pattern ofimproved chondrogenesis both in micro-pellets and at2% O2 was previously observed in hBMSC chondrogeniccultures [3]. For both oBMSC (in this study) andhBMSC (previous studies [3, 7]), more homogeneous tis-sue was generated using micro-pellet cultures, relative to

macro-pellet cultures. The effect of atmospheric O2 con-centration was more discernible in homogeneous micro-pellet models, demonstrating that, as with hBMSC chon-drogenesis, oBMSC chondrogenesis appears to be moreefficient at lower O2 concentrations. Across all oBMSCdonors, the greatest GAG/DNA was observed in micro-pellets, compared to macro-pellets, at optimal condi-tions. These data demonstrate that the per-cell matrixoutput is maximised in the micro-pellet model.Next, we sought to evaluate the use of autologous

oBMSC micro-pellets in an in vivo sheep pilot study.Since autologous ACh are already being used in variousclinical cartilage repair strategies [35], we included ex-panded oACh micro-pellets for comparison. A 10-daychondrogenic induction culture, rather than 14 days, wasselected based on a previous study that indicated thatfollowing ~ 10 days of culture, micro-pellets gained sig-nificant mass and yet still retained their propensity toamalgamate into a continuous tissue [10]. We charac-terised oBMSC and oACh micro-pellets from 2 uniquedonors each, using histology and qPCR. By day 10, bothoBMSC and oACh micro-pellets stained for GAG withAlcian blue, and both tissues exhibited similar tissuemorphology. Conversely, qPCR analysis revealed thathypertrophic gene expression (COL10A1 and RUNX2)was absent in oACh micro-pellets, but significantly over-represented in oBMSC micro-pellets. Hypertrophic sig-nalling and upregulation of endochondral ossificationpathways is a well-characterised problem in hBMSCchondrogenic cultures and is viewed as a major impedi-ment to BMSC-mediated cartilage repair [7, 36]. It iscommon to demonstrate hBMSC hypertrophy byimplanting tissue ectopically in immune compromisedmice [3, 7, 22, 37]. However, it is unclear if the highlyvascularised ectopic microenvironment exacerbates thepropensity of hBMSC to undergo hypertrophy, and ifhypertrophy would be less problematic in the less vascu-lar environment found in a synovial joint. For this rea-son, there is merit in evaluating the capacity of BMSC toregenerate cartilage in actual cartilage defect models.To assess the potential of micro-pellets to repair

osteochondral defects, we performed a large animal pilotstudy using sheep and micro-pellets formed from autolo-gous cells. Two sheep were assigned to the oBMSCgroup and 2 sheep were assigned to the oACh group.Three replicate defects were created in one stifle jointon each animal. All defects were 6 mm in diameter and1.5 mm deep, penetrating through the cartilage and intothe subchondral bone. Micro-pellets were packed intocylindrical moulds, bonded to a Celgro™ collagen mem-brane with fibrin glue, and then anchored into defectsites with fibrin glue and sutures (see Fig. 5). Duringpost-surgical recovery, slings were used to partially re-duce weight-bearing on the treated sheep joints and


plaster casts were introduced to prevent joint mobility.Our ethics protocol did not allow for completeunweighting of the repaired joint. In clinical cartilage re-pair procedures such as MACI (reviewed here [38]), hu-man patients are advised to adhere to strict progressiveweight-bearing and passive motion following these pro-cedures. As the sheep knee joint is high up on the ani-mal’s flank, it is challenging to immobilise this joint withwrapping or a cast. Despite the formation of three 6-mmdiameter osteochondral defects in one stifle joint, ani-mals did not appear to be in pain or appear to reduceload on the repaired leg while in the sling or after releasefrom the sling. Full unloading of the treated joint for1 week or more could potentially improve engraftmentand the stability of repair tissues.Following 8 weeks of in vivo incubation, defects filled

with oBMSC micro-pellets were not effectively repaired.oBMSC micro-pellets were visible in 1 of 5 defect sec-tions, but the lacunae structure and Alcian blue stainingsuggested that the cartilage-like tissue was either de-differentiating or being absorbed by surrounding tissue.H&E staining suggested that most of the defect fill vol-ume contained fibrous tissue, rather than cartilage-likerepair tissue. A layer of tissue was stained faintly withAlcian blue at the base of some defect sites, but it is pos-sible that this tissue was derived from adjacent nativetissue, or residual oBMSC micro-pellets, rather than newcartilage tissue. Repair from adjacent tissue might beconsistent with the concept that BMSC-based therapiescan upregulate endogenous regenerative processes [39].However, this repair was modest and our limited con-trols did not allow for a statistical evaluation of this out-come. In previous studies, we implanted hBMSC micro-pellets subcutaneously in NSG mice, and after 8 weeks,we observed mineralisation and the formation ofmarrow-like structures [3]. This is consistent with a widebody of literature that suggests that chondrogenically in-duced hBMSC appear to engage intrinsic endochondralossification programming, triggering vascular penetra-tion, ossification, and support of marrow formation [36,40]. However, in this study, we observed no evidence ofmineralisation or marrow formation from oBMSCmicro-pellets implanted into osteochondral defect sites.Another study that implanted oBMSC in TGF-β1-ladengelatin foams in sheep tibial growth plates similarly re-ported a lack of mineralised tissue, with mostly fibroustissue present after 5 weeks [41]. It is possible that dur-ing the incubation period in our studies, micro-pelletmineralisation occurred, but that this tissue was com-pletely re-absorbed by immune cells. Alternatively, it ispossible that the microenvironment within these jointdefect sites might mitigate mineralisation of tissue, rela-tive to more vascularised microenvironments such as asubcutaneous mouse pouch. While we did not see

evidence of mineralisation at 8 weeks, oBMSC micro-pellets did not yield promising cartilage-like repairtissue.In the oACh micro-pellet group, evidence of cartilage

repair tissue was visible in 2 of 5 histology sections.While there was evidence of continuous cartilage-liketissue formation in these regions, there were also regionsof fibrous tissue and vasculature. Where cartilage repairwas visible, this repair tissue filled both the articular car-tilage defect and the subchondral bone region. Maturelacunae structures and Alcian blue staining were visiblein this repair tissue. Longer incubation periods beyond8 weeks would be required to determine if repair tissueintegration and a high-quality cartilage repair tissuemight evolve. Many sheep studies are carried out forlonger periods (8, 10, and 12 weeks [42]; 3 months [43];4–12 months [44]), and we presume that promising as-pects of repair with oACh micro-pellets would have im-proved with time. The low success rate of establishing arepair tissue in the oACh group may also be improvedby further reducing weight-bearing and motion of thejoints during the first few weeks of recovery.The success of BMSC in cartilage repair is dependent

on the development of strategies that promote the for-mation of stable articular cartilage and mitigate hyper-trophy. In recent work, we characterised hBMSCchondrogenic and hypertrophic differentiation cascadesin response to TGF-β1 [7]. We observed that in re-sponse to as little as a single day of TGF-β1 exposure,intrinsic hypertrophic signalling in hBMSC is activated.SP7, a transcription factor that drives bone formationand hypertrophy [45, 46], was upregulated following thisbrief TGF-β1 exposure and persisted even when TGF-β1was withdrawn. Knockdown of Sp7 in mice reduceshypertrophy [45], and SP7 would be a logical drug orgenetic engineering target to prevent hypertrophy. Otherlogical targets to obstruct hypertrophy include BMP2and WNT11 pathways, which have been trialled withsome success in previous hBMSC studies [47, 48]. Cur-rently, chondrogenic stability and the absence of hyper-trophy remain significant obstacles for the use of BMSC-derived tissues in treating cartilage defect repair, andthese obstacles need to be resolved definitively prior toclinical application [7, 36]. The broad efforts to developnew technologies and clinical trials to repair cartilageusing cell-based products are reviewed here [49–51].

ConclusionIn summary, cell-mediated cartilage repair strategies inthe leading large animal cartilage repair model, thesheep, remain in their infancy. Identifying optimaloBMSC culture conditions and understanding the limi-tations of oBMSC will likely be critical to advancing thefield of cartilage defect repair. Our data demonstrate


that (1) there is limited cross-reactivity between com-mon hBMSC antibody panels and oBMSC, but thatCD44, CD146, and CD45 are useful markers; (2) IBMXis toxic during oBMSC adipogenic culture, but IBMX isnot required and adipogenic induction can be salvagedby excluding this molecule from the medium; (3)oBMSC chondrogenesis can be enhanced by usingmicro-pellet cultures and reduced oxygen atmospheresin vitro; (4) micro-pellets formed from oBMSC appearmorphologically similar to micro-pellets formed fromoACh, but have upregulated hypertrophic gene expres-sion in vitro; and (5) oBMSC micro-pellets produce in-ferior cartilage repair tissue in vivo, compared to oAChmicro-pellets.

Supplementary informationThe online version contains supplementary material available at https://doi.org/10.1186/s13287-020-02045-3.

Additional file 1: Supplementary Figure 1. Flow cytometry analysisof cell surface markers for oBMSC 1.



Additional file 4: Supplementary Figure 4. Control flow cytometryanalysis of cell surface markers for fresh sheep mononuclear cells (MNC)isolated from sheep bone marrow aspirate.

Additional file 5: Supplementary Figure 5. Control flow cytometryanalysis of cell surface markers on expanded hBMSC.

Additional file 6: Supplementary Figure 6. Microscopic images overa 14-day culture period of two additional oBMSC donors cultured asmacro-pellets in deep-well plates (A and C), and cultured as micro-pelletsin the Microwell-mesh platform (B and D). Scale bar = 1 mm.

Additional file 7: Supplementary Figure 7. Type II collagen stainingin A) macro-pellet and B) micro-pellet sections. Scale bar = 400 μm.

Additional file 8: Supplementary Figure 8. Type I collagen staining inA) macro-pellet and B) micro-pellet sections. Scale bar = 400 μm.

Additional file 9: Supplementary Figure 9. Type X collagen stainingin A) macro-pellet and B) micro-pellet sections. Scale bar = 400 μm.

Additional file 10: Supplementary Figure 10. Scaffold-only controlsfor A) oACh sheep pilot, and B) oBMSC sheep pilot. The middle and rightimages are enlarged from the boxes indicated in the images on the left.The * points to non-degraded scaffold, and the triangle points to a bloodvessel. Tissue was fibrous with negligible repair.

Additional file 11: Supplementary Figure 11. Photos of joints 8weeks after surgery. Asterisk indicates scaffold-only controls. Arrow pointsto the best repair observed with oACh micro-pellets.

AbbreviationsACAN: Aggrecan; ANOVA: Analysis of variance; Anti-Anti: Antibiotic-Antimycotic; BMP2: Bone Morphogenetic Protein 2; BMSC: Bone marrowstromal cells; BSA: Bovine serum albumin; COL10A1: Collagen Type X Alpha 1Chain; COL1A1: Collagen Type I Alpha 1 Chain; COL2A1: Collagen Type IIAlpha 1 Chain; Dex: Dexamethasone; DMSO: Dimethyl sulfoxide;DNA: Deoxyribonucleic acid; dsDNA: Double stranded deoxyribonucleic acid;ECM: Extracellular matrix; EDTA: Ethylenediaminetetraacetic acid;GAG: Glycosaminoglycan; GAPDH: Glyceraldehyde 3-phosphate dehydrogen-ase; H&E: Haematoxylin and eosin; hBMSC: Human bone marrow stromalcells; HG-DMEM: High glucose Dulbecco’s modified Eagle’s medium; IBMX: 3-Isobutyl-1-methylxanthine; LG-DMEM: Low glucose Dulbecco’s modifiedEagle’s medium; MNC: Mononuclear cells; oACH: Ovine articular

chondrocytes; oBMSC: Ovine bone marrow stromal cells; PBS: Phosphate-buffered saline; PDMS: Polydimethylsiloxane; PenStrep: Penicillin/streptomycin; PFA: Paraformaldehyde; qPCR: Quantitative polymerase chainreaction; QUT: Queensland University of Technology; RNase: Ribonuclease;RUNX2: RUNX Family Transcription Factor 2; SOX9: SRY-Box TranscriptionFactor 9; WNT11: Wnt family member 11

AcknowledgementsThe Translational Research Institute (TRI) is supported by TherapeuticInnovation Australia (TIA). TIA is supported by the Australian Governmentthrough the National Collaborative Research Infrastructure Strategy (NCRIS)programme. The authors thank the core facilities at the TRI, including theFlow Cytometry Core and the Microscopy Core. The authors thank Dr.Siamak Saifzadeh’s team at the Medical Engineering Research Facility (MERF,QUT) for assistance in organising and executing this study. The authorswould like to thank Professor MingHao Zheng for technical advice, andOrthocell Ltd. (Perth Australia) for donating the Celgro scaffolds used inthese studies. MRD, RWC, and TJK gratefully acknowledge project supportfrom the National Health and Medicine Research Council (NHMRC) ofAustralia (Project Grant APP1083857) and NHMRC Fellowship support of MRD(APP1130013). KF and PGR are supported by the Intramural ResearchProgram of the NIH, NIDCR.

Authors’ contributionsKF, EM, PGR, SG, RWC, SS, TJK, and MRD designed the research, analysed thedata, and wrote the paper; EM, KF, SS, and MRD performed the research. Theauthors read and approved the final manuscript.

FundingMRD, RWC, and TJK gratefully acknowledge project support from theNational Health and Medicine Research Council (NHMRC) of Australia (ProjectGrant APP1083857) and NHMRC Fellowship support of MRD (APP1130013).KF and PGR are supported by the Intramural Research Program of the NIH,NIDCR. The funding bodies played no role in the design of the study andcollection, analysis, and interpretation of data or writing of the manuscript.

Availability of data and materialsAll data is available through the senior author.

Ethics approval and consent to participateHuman bone marrow was collected from informed consenting healthyvolunteer donors. Donation and use of cells were approved by the MaterHospital Human Research Ethics Committee and the Queensland Universityof Technology Human Research Ethics Committee. All animal procedureswere approved by the Queensland University of Technology Animal EthicsCommittee.

Consent for publicationNot applicable.

Competing interestsCo-authors have no conflict of interest to report. Authors, their immediatefamily, and any research foundation to which they might be affiliated to didnot receive any financial payments or other benefits from any commercialentity related to the subject of this article.

Author details1Centre for Biomedical Technologies (CBT), Queensland University ofTechnology (QUT), Brisbane, Queensland, Australia. 2National Institute ofDental and Craniofacial Research (NIDCR), National Institutes of Health (NIH),Bethesda, Maryland, USA. 3Translational Research Institute (TRI), Brisbane,Queensland, Australia. 4School of Biomedical Sciences, Queensland Universityof Technology (QUT), Brisbane, Queensland, Australia. 5Adelaide MedicalSchool, Faculty of Health and Medical Sciences, University of Adelaide,Adelaide, South Australia, Australia. 6Mater Research Institute – University ofQueensland (UQ), Translational Research Institute (TRI), Brisbane, Queensland,Australia.


https://doi.org/10.1186/s13287-020-02045-3https://doi.org/10.1186/s13287-020-02045-3

Received: 25 July 2020 Accepted: 23 November 2020

References1. Negoro T, Takagaki Y, Okura H, Matsuyama A. Trends in clinical trials for

articular cartilage repair by cell therapy. NPJ Regen Med. 2018;3:17.2. Serafini M, Sacchetti B, Pievani A, Redaelli D, Remoli C, Biondi A, Riminucci

M, Bianco P. Establishment of bone marrow and hematopoietic nichesin vivo by reversion of chondrocyte differentiation of human bone marrowstromal cells. Stem Cell Res. 2014;12(3):659–72.

3. Futrega K, Palmer JS, Kinney M, Lott WB, Ungrin MD, Zandstra PW,Doran MR. The microwell-mesh: a novel device and protocol for thehigh throughput manufacturing of cartilage microtissues. Biomaterials.2015;62:1–12.

4. Cook JL, Hung CT, Kuroki K, Stoker AM, Cook CR, Pfeiffer FM, ShermanSL, Stannard JP. Animal models of cartilage repair. Bone Joint Res. 2014;3(4):89–94.

5. Moran CJ, Ramesh A, Brama PA, O'Byrne JM, O'Brien FJ, Levingstone TJ. Thebenefits and limitations of animal models for translational research incartilage repair. J Exp Orthop. 2016;3(1):1.

6. Music E, Futrega K, Doran MR. Sheep as a model for evaluatingmesenchymal stem/stromal cell (MSC)-based chondral defect repair.Osteoarthr Cartil. 2018;26(6):730–40.

7. Futrega K, Robey PG, Klein TJ, Crawford RW, Doran MR. Micro-pellet culturereveals that bone marrow mesenchymal stromal cell (BMSC) chondrogenicinduction is triggered by a single day of TGF-beta1 exposure. bioRxiv. 2019.https://doi.org/10.1101/853556.

8. Markway BD, Tan GK, Brooke G, Hudson JE, Cooper-White JJ, Doran MR.Enhanced chondrogenic differentiation of human bone marrow-derivedmesenchymal stem cells in low oxygen environment micropellet cultures.Cell Transplant. 2010;19(1):29–42.

9. Johnstone B, Hering TM, Caplan AI, Goldberg VM, Yoo JU. In vitrochondrogenesis of bone marrow-derived mesenchymal progenitor cells.Exp Cell Res. 1998;238(1):265–72.

10. Babur BK, Futrega K, Lott WB, Klein TJ, Cooper-White J, Doran MR. High-throughput bone and cartilage micropellet manufacture, followed byassembly of micropellets into biphasic osteochondral tissue. Cell Tissue Res.2015;361(3):755–68.

11. Murphy CL, Thoms BL, Vaghjiani RJ, Lafont JE. Hypoxia. HIF-mediatedarticular chondrocyte function: prospects for cartilage repair. Arthritis ResTher. 2009;11(1):213.

12. Thoms BL, Dudek KA, Lafont JE, Murphy CL. Hypoxia promotes theproduction and inhibits the destruction of human articular cartilage.Arthritis Rheum. 2013;65(5):1302–12.

13. Robins JC, Akeno N, Mukherjee A, Dalal RR, Aronow BJ, Koopman P,Clemens TL. Hypoxia induces chondrocyte-specific gene expression inmesenchymal cells in association with transcriptional activation of Sox9.Bone. 2005;37(3):313–22.

14. Berner A, Henkel J, Woodruff MA, Saifzadeh S, Kirby G, Zaiss S, Gohlke J,Reichert JC, Nerlich M, Schuetz MA, Hutmacher DW. Scaffold-cell boneengineering in a validated preclinical animal model: precursors vsdifferentiated cell source. J Tissue Eng Regen Med. 2017;11(7):2081–9.

15. Ungrin MD, Joshi C, Nica A, Bauwens C, Zandstra PW. Reproducible, ultrahigh-throughput formation of multicellular organization from single cellsuspension-derived human embryonic stem cell aggregates. PLoS One.2008;3(2):e1565.

16. Chambers KF, Mosaad EM, Russell PJ, Clements JA, Doran MR. 3D cultures ofprostate cancer cells cultured in a novel high-throughput culture platformare more resistant to chemotherapeutics compared to cells cultured inmonolayer. PLoS One. 2014;9(11):e111029.

17. Bianco P, Robey PG, Saggio I, Riminucci M. “Mesenchymal” stem cells inhuman bone marrow (skeletal stem cells): a critical discussion of theirnature, identity, and significance in incurable skeletal disease. Hum GeneTher. 2010;21(9):1057–66.

18. Deschaseaux F, Gindraux F, Saadi R, Obert L, Chalmers D, Herve P. Directselection of human bone marrow mesenchymal stem cells using an anti-CD49a antibody reveals their CD45med,low phenotype. Br J Haematol.2003;122(3):506–17.

19. Siddappa R, Licht R, van Blitterswijk C, de Boer J. Donor variation and loss ofmultipotency during in vitro expansion of human mesenchymal stem cellsfor bone tissue engineering. J Orthop Res. 2007;25(8):1029–41.

20. Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini F, Krause D,Deans R, Keating A, Prockop D, Horwitz E. Minimal criteria for definingmultipotent mesenchymal stromal cells. The International Society forCellular Therapy position statement. Cytotherapy. 2006;8(4):315–7.

21. Sorrentino A, Ferracin M, Castelli G, Biffoni M, Tomaselli G, Baiocchi M, FaticaA, Negrini M, Peschle C, Valtieri M. Isolation and characterization of CD146+multipotent mesenchymal stromal cells. Exp Hematol. 2008;36(8):1035–46.

22. Sacchetti B, Funari A, Michienzi S, Di Cesare S, Piersanti S, Saggio I, TagliaficoE, Ferrari S, Robey PG, Riminucci M, Bianco P. Self-renewingosteoprogenitors in bone marrow sinusoids can organize a hematopoieticmicroenvironment. Cell. 2007;131(2):324–36.

23. Shi S, Gronthos S. Perivascular niche of postnatal mesenchymal stem cells inhuman bone marrow and dental pulp. J Bone Miner Res. 2003;18(4):696–704.

24. Gronthos S, Simmons PJ. The growth factor requirements of STRO-1-positive human bone marrow stromal precursors under serum-deprivedconditions in vitro. Blood. 1995;85(4):929–40.

25. Petters O, Schmidt C, Thuemmler C, Peinemann F, Zscharnack M, SomersonJS, Schulz RM. Point-of-care treatment of focal cartilage defects withselected chondrogenic mesenchymal stromal cells - an in vitro proof-of-concept study. J Tissue Eng Regen Med. 2018;12(7):1717–27.

26. Barilani M, Banfi F, Sironi S, Ragni E, Guillaumin S, Polveraccio F, Rosso L,Moro M, Astori G, Pozzobon M, Lazzari L. Low-affinity nerve growth factorreceptor (CD271) heterogeneous expression in adult and fetal mesenchymalstromal cells. Sci Rep. 2018;8(1):9321.

27. Sanjurjo-Rodriguez C, Castro-Vinuelas R, Hermida-Gomez T, Fernandez-Vazquez T, Fuentes-Boquete IM, de Toro-Santos FJ, Diaz-Prado SM, Blanco-Garcia FJ. Ovine mesenchymal stromal cells: morphologic, phenotypic andfunctional characterization for osteochondral tissue engineering. PLoS One.2017;12(1):e0171231.

28. McCarty RC, Gronthos S, Zannettino AC, Foster BK, Xian CJ. Characterisationand developmental potential of ovine bone marrow derived mesenchymalstem cells. J Cell Physiol. 2009;219(2):324–33.

29. Essayan DM. Cyclic nucleotide phosphodiesterases. J Allergy Clin Immunol.2001;108(5):671–80.

30. Scott MA, Nguyen VT, Levi B, James AW. Current methods of adipogenicdifferentiation of mesenchymal stem cells. Stem Cells Dev. 2011;20(10):1793–804.

31. Rentsch C, Hess R, Rentsch B, Hofmann A, Manthey S, Scharnweber D,Biewener A, Zwipp H. Ovine bone marrow mesenchymal stem cells: isolationand characterization of the cells and their osteogenic differentiation potentialon embroidered and surface-modified polycaprolactone-co-lactide scaffolds. InVitro Cell Dev Biol Anim. 2010;46(7):624–34.

32. Sales VL, Mettler BA, Lopez-Ilasaca M, Johnson JA Jr, Mayer JE Jr.Endothelial progenitor and mesenchymal stem cell-derived cells persistin tissue-engineered patch in vivo: application of green and redfluorescent protein-expressing retroviral vector. Tissue Eng. 2007;13(3):525–35.

33. Krinner A, Zscharnack M, Bader A, Drasdo D, Galle J. Impact of oxygenenvironment on mesenchymal stem cell expansion and chondrogenicdifferentiation. Cell Prolif. 2009;42(4):471–84.

34. Zscharnack M, Poesel C, Galle J, Bader A. Low oxygen expansionimproves subsequent chondrogenesis of ovine bone-marrow-derivedmesenchymal stem cells in collagen type I hydrogel. Cells TissuesOrgans. 2009;190(2):81–93.

35. Armoiry X, Cummins E, Connock M, Metcalfe A, Royle P, Johnston R,Rodrigues J, Waugh N, Mistry H. Autologous chondrocyte implantation withchondrosphere for treating articular cartilage defects in the knee: anevidence review group perspective of a NICE single technology appraisal.Pharmacoeconomics. 2019;37(7):879–86.

36. Somoza RA, Welter JF, Correa D, Caplan AI. Chondrogenic differentiation ofmesenchymal stem cells: challenges and unfulfilled expectations. Tissue EngPart B Rev. 2014;20(6):596–608.

37. Robey PG, Kuznetsov SA, Riminucci M, Bianco P. Bone marrow stromal cellassays: in vitro and in vivo. Methods Mol Biol. 2014;1130:279–93.

38. Edwards PK, Ackland T, Ebert JR. Clinical rehabilitation guidelines for matrix-induced autologous chondrocyte implantation on the tibiofemoral joint. JOrthop Sports Phys Ther. 2014;44(2):102–19.

39. Gupta PK, Das AK, Chullikana A, Majumdar AS. Mesenchymal stem cells forcartilage repair in osteoarthritis. Stem Cell Res Ther. 2012;3(4):25. https://doi.org/10.1186/scrt116.


https://doi.org/10.1101/853556https://doi.org/10.1186/scrt116https://doi.org/10.1186/scrt116

40. Bianco P, Robey PG, Simmons PJ. Mesenchymal stem cells: revisiting history,concepts, and assays. Cell Stem Cell. 2008;2(4):313–9.

41. McCarty RC, Xian CJ, Gronthos S, Zannettino AC, Foster BK. Application ofautologous bone marrow derived mesenchymal stem cells to an ovinemodel of growth plate cartilage injury. Open Orthop J. 2010;4:204–10.

42. Jones CW, Willers C, Keogh A, Smolinski D, Fick D, Yates PJ, Kirk TB, ZhengMH. Matrix-induced autologous chondrocyte implantation in sheep:objective assessments including confocal arthroscopy. J Orthop Res. 2008;26(3):292–303.

43. Bozkurt M, Aşık MD, Gürsoy S, Türk M, Karahan S, Gümüşkaya B, Akkaya M,Şimşek ME, Cay N, Doğan M. Autologous stem cell-derived chondrocyteimplantation with bio-targeted microspheres for the treatment ofosteochondral defects. Orthop Surg Res. 2019;14:394. https://doi.org/10.1186/s13018-019-1434-0.

44. Dorotka R, Bindreiter U, Macfelda K, Windberger U, Nehrer S. Marrowstimulation and chondrocyte transplantation using a collagen matrix forcartilage repair. Osteoarthr Cartil. 2005;13(8):655–64.

45. Kaback LA, Soung do Y, Naik A, Smith N, Schwarz EM, O'Keefe RJ, Drissi H.Osterix/Sp7 regulates mesenchymal stem cell mediated endochondralossification. J Cell Physiol. 2008;214(1):173–82.

46. Rashid H, Ma C, Chen H, Wang H, Hassan MQ, Sinha K, de Crombrugghe B,Javed A. Sp7 and Runx2 molecular complex synergistically regulateexpression of target genes. Connect Tissue Res. 2014;55(Suppl 1):83–7.

47. Occhetta P, Pigeot S, Rasponi M, Dasen B, Mehrkens A, Ullrich T, Kramer I,Guth-Gundel S, Barbero A, Martin I. Developmentally inspired programmingof adult human mesenchymal stromal cells toward stable chondrogenesis.Proc Natl Acad Sci U S A. 2018;115(18):4625–30.

48. Diederichs S, Tonnier V, Marz M, Dreher SI, Geisbusch A, Richter W.Regulation of WNT5A and WNT11 during MSC in vitro chondrogenesis:WNT inhibition lowers BMP and hedgehog activity, and reduceshypertrophy. Cell Mol Life Sci. 2019;76(19):3875–89.

49. Arshi A, Petrigliano FA, Williams RJ, Jones KJ. Stem cell treatment for kneearticular cartilage defects and osteoarthritis. Curr Rev Musculoskelet Med.2020;13(1):20–7.

50. Lam ATL, Reuveny S, Oh SK-W. Human mesenchymal stem cell therapy forcartilage repair: review on isolation, expansion, and constructs. Stem CellRes. 2020;44:101738. https://doi.org/10.1016/j.scr.2020.101738.

51. Eder C, Schmidt-Bleek K, Geissler S, Sass FA, Maleitzke T, Pumberger M,Perka C, Duda GN, Winkler T. Mesenchymal stromal cell and bone marrowconcentrate therapies for musculoskeletal indications: a concise review ofcurrent literature. Mol Biol Rep. 2020;47(6):4789–814.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.


https://doi.org/10.1186/s13018-019-1434-0https://doi.org/10.1186/s13018-019-1434-0https://doi.org/10.1016/j.scr.2020.101738

Outline placeholderAbstractObjectiveDesignResultsConclusion

IntroductionMaterials and methodsLive animal anaesthesia and non-surgical proceduresBone marrow aspiration and BMSC cultureFlow cytometry characterisationArticular cartilage biopsy and oACh cultureFabrication of the Microwell-mesh culture deviceOsteogenic and adipogenic inductionChondrogenic induction culturesQuantification of glycosaminoglycans (GAG) and DNAQuantitative PCR (qPCR)Histology and immunohistochemistryAutologous repair of invivo cartilage defects in sheepStatistical analysis

ResultsoBMSC characterisationGrowth of macro-pellets and micro-pellets in the microwell-mesh systemLow oxygen increases GAG production in micro-pelletsECM characterisation in macro-pellets and micro-pelletsIn vivo cartilage defects in sheepHistological and qPCR assessment of oBMSC and oACh micro-pellets pre-implantHistological assessment of oBMSC and oACh micro-pellets post-implant

DiscussionConclusionSupplementary informationAbbreviationsAcknowledgementsAuthors’ contributionsFundingAvailability of data and materialsEthics approval and consent to participateConsent for publicationCompeting interestsAuthor detailsReferencesPublisher’s Note

Date post:	26-Jan-2021
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

Characterisation of ovine bone marrow-derived stromal ... · oBMSC micro-pellets, while another two...

Documents